Theta and alpha oscillations during sleep predict subsequent dream recall.

-last years, the underlying brain mechanism correlates are yet largely unknown. Indeed, dream contents are accessible only indirectly through the verbal report collected after awakenings; hence, a limited number of studies investigated the EEG features preceding dream recall. Objectives: The aim of the study was to evaluate if the presence/absence of dream recall may be associated to specific topographical features of the EEG preceding the awakening. Methods: 64 subjects (34 M; mean age= 23.6±2.1 yrs) were divided in 4 different groups: subjects who recalled (REC) and non-recalled (NREC) dreams after both REM and stage 2 NREM morning awakenings. Dreams were collected by a diary compiled upon morning awakening from REM and stage 2. The 5 min of sleep preceding the awakening were recorded by 19 cortical derivations, and were analysed across these EEG bands: delta (0.5 4.75 Hz), theta (5 7.75 Hz), alpha (8 11.75 Hz), sigma (12 14.75 Hz), and beta (15 24.75 Hz). Statistical comparisons have been carried out between REC and NREC groups, separately for the two sleep stages. Results: Analyses revealed the presence of higher theta power in REC than in NREC group during REM sleep in frontal areas. During stage 2, the alpha activity in right temporo-parietal areas was lower in REC than in NREC group. A correlational analysis showed that actual number of dreams recalled upon awakening from REM sleep and from stage 2 was respectively predicted by the extent of theta and alpha power, with a regional specificity. Conclusions: The current study finds electrophysiological and topographical substrate of the presence/absence of dream recall during REM and NREM sleep. Higher frontal theta and lower right temporo-parietal alpha activity showed a state-specific association with successful dream recall. Therefore, our findings also provide an electrophysiological and topographic support to the two generator models of dream experience.